Analysis (Ie)
g of Blobs (I3)
sult
4.5 Further Steps
Future analyzing of the type of hatched patterns (diagonal
or parallel) will enable the determination of the use of build-
ings (residential buildings or outbuildings) besides the pure
location information.
Further steps will be the elimination of remaining errors as
well as the hatching lines. This way the final vectorization
only will have to deal with the empty outlines of the buildings.
Concerning the building outlines, it is planned not only to ex-
tract the outer boundaries (transition of blobs to background)
but also to extract the inner boundaries between adjacent
houses. However, the latter task will be more complicated.
Furthermore, the addition of two more analysis directions,
the diagonals, in terms of the investigation of the runlength
encoded vectors is planned. That will probably reduce ambi-
guities and make the results more robust.
Additionally, it has to be thought of relating the empirically
found dimensions of the structuring elements to the scanning
resolution and to different line widths. Possibilities of includ-
ing algorithms for assessing the obtained results should be
included.
5 CONCLUSIONS
The paper demonstrates a procedure for a raster based, au-
tomated recognition process of buildings within the german
base map 1:5000. The identification of the buildings includes
the recognition of the hatched filling patterns as well as the
outlines of the buildings. Although the results represent a
preliminary stage, they show the general potential of raster
based techniques, such as mathematical morphology, for au-
tomated map understanding tasks.
The advantages of the introduced raster based approach can
be summarized in the following way:
— low complexity of the approach
— maintenance of the positional accuracy because of the
acting on the original image data up to an advanced
processing stage
— strongly reduced amount of vector data.
Since an enormous amount of digital building data has to be
captured, e.g. for the creation of 3D models of towns, the
proposed approach is a useful tool to obtain the 2D outlines
of buildings as vector representations in an efficient way.
REFERENCES
Ablameyko S. et al., 1993. A System for Automatic Vectoriza-
tion and Interpretation of Map-drawings, Proceedings 4.
ICCV, Berlin, IEEE Computer Society Press, Los Alami-
tos, California, pp. 456-460.
Antoine D., Collin S., Tombre K., 1992. Analysis of Techni-
cal Documents: The REDRAW System, in: Structured
Document Image Analysis, Baird/Bunke/Yamamoto
(Eds.), Springer, pp. 385-402.
Boatto L., Consorti V., Del Buono M., Di Zenzo S. et al.,
1992. An Interpretation System for Land Register Maps,
IEEE Computer Magazine, 25(7), pp. 25-33.
Crosilla F., Piccinini L.C., 1991. A new approach for pat-
tern recognition in cadastral cartography, Geoinforma-
tion Systems (GIS), 4/91, Wichmann Verlag, Karlsruhe,
pp. 14-20
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996
Ebi N., 1993. Eine Methodik zur frame-basierten Mod-
ellierung und Bildinterpretation von topographischen
Karten mittleren MaBstabs, Bremer Beitráge zur Ge-
ographie und Raumplanung, Heft 26, Bahrenberg G. et
al. (Eds.), Universitát Bremen, Dissertation
Haralick R.M., Shapiro L.G., 1992. Computer and Robot Vi-
sion, Addison-Wesley, Vol.1.
Hori O., Okazaki A., 1992. High Quality Vectorization Based
on a Generic Object Model, in: Structured Docu-
ment Image Analysis, Baird/Bunke/Yamamoto (Eds.),
Springer, pp. 325-339.
lllert A., 1990. Automatische Erfassung von Karten-
schrift, Symbolen und GrundriBobjekten aus der
Deutschen Grundkarte 1:5000, Wissenschaftliche Ar-
beit der Fachrichtung Vermessungswesen der Universitat
Hannover, Nr. 166.
Kasturi R. et al., 1990. A system for interpretation of line
drawings, IEEE Transactions on PAMI, Vol. 12, No. 10,
pp. 978-992.
Klauer R., 1993. Untersuchungen zur Optimierung von
Verfahren der automatisierten Digitalisierung von
Flurkarten, Wissenschaftliche Arbeit der Fachrichtung
Vermessungswesen der Universität Hannover, Nr. 183.
Matheron G., 1975. Random Sets and Integral Geometry, Wi-
ley, New York.
Mayer H., 1994. Automatische wissensbasierte Extraktion von
semantischer Information aus gescannten Karten, Dis-
sertation am Lehrstuhl für Photogrammetrie und Fern-
erkundung, Deutsche Geodätische Kommission, Reihe
C, Heft Nr. 417, München.
Quint F., Bähr H.P., 1994. Feature extraction for map based
image interpretation, Proceedings of the Third Interna-
tional Colloquium LIESMARS: Integration, Automation
and Intelligence in Photogrammetry, Remote Sensing
and GIS, Shi/Du/Gao (Eds.), Wuhan, China, pp. 1-8.
Quint F., Sties M., 1995. Map-based semantic modeling for
the extraction of objects from aerial images, in: Auto-
matic Extraction of man-made Objects from Aerial and
Space Images, Grün/Kübler/Agouris (Eds.), Birkhäuser,
Basel, pp. 307-316
Serra J., 1982. Image Analysis and Mathematical Morphol-
ogy, Academic Press, London.
Shen W., Ebi N., Besslich P., 1991. Schraffurerkennung in
graphischen Vorlagen, Mustererkennung 91, 13. DAGM-
Symposium, Springer Verlag, pp. 407-412.
Stengele R., 1995. Kartographische Mustererkennung —
Rasterorientierte Verfahren zur Erfassung von Geo-
Informationen, Dissertation ETH Nr. 11099, Institut für
Geodäsie und Photogrammetrie an der ETH Zürich, Nr.
54.
Suzuki S., Yamada T., 1990. MARIS: Map Recognition In-
put System, NATO ASI Series, Vol. F 65, Mapping and
Spatial Modelling for Navigation, Springer Verlag, pp.
95-116.
Yamada H., Yamamoto K., 1993. Directional mathematical
morphology and reformalized Hough Transformation for
the analysis of topographic maps, IEEE Transactions on
Pattern Analysis and Machine Intelligence, Vol. 15, No.
4, pp. 380-387.
87
